Creating a Gene Regulatory Network Comparing a Wild Type Strain with a Mutant ΔGLN3 Deletion in S....

Creating a Gene Regulatory Network Comparing a Wild Type Strain with a

Mutant ΔGLN3 Deletion in S. cerevisiae Showed that ΔGLN3 Exhibits No Meaningful Control of Cold Shock

Response

Alyssa N. Gomes & Tessa A. Morris

Loyola Marymount University

Biomathematical Modeling

Dr. Dahlquist & Fitzpatrick

May 7, 2015

Understand the Effect of Deleting gene, ΔGLN3, on the Cell’s Cold Shock Response

• What transcription factors in S. cerevisiae control the cold shock response responsible for maintaining homeostasis?• Discuss the importance of cold shock, microarray analysis, ΔGLN3• Statistical analysis leading to profile selection

• Choosing profile 45 to create a gene regulatory network• Statistical Analysis: p-values, Bonferroni, B-H tests• Clustering and Gene Oncology Analysis with STEM: STEM analysis, GO

terms, significant profiles• Use YEASTRACT and GRNsight to determine which transcription factors

regulate a cluster of genes and that “ONLY DNA binding” should be pursued further

• Run model one where the threshold parameters b are not estimated and one where they are estimated

• ΔGLN3 exhibits no meaningful control of cold shock response, other more connected genes should be studied.

What Transcription Factors in S. cerevisiae Control the Cold Shock Response Responsible for Maintaining Homeostasis?

• ΔGLN3 was noted as a transcriptional activator of genes regulated by nitrogen catabolite repression and a lack of nitrogen sources

• Cold Shock, the response to a sudden decline in temperature, is not well studied.

• DNA microarray analysis was performed to analyze the gene expression thousands of genes at once• Red spot if amount of RNA expressed is increased • Green spot if decreased

• Log fold change was calculated for time points (15, 30, 60, 90, 120)

• Conducted with cold shock at 13°C until 60m, then recovery 30°C

Choosing Profile 45 to Create a Gene Regulatory Network From

• Statistical Analysis: • Determine the number of genes that have a log fold

change that is different than zero at one or more time points

• p-values, Bonferroni, B-H tests• Clustering and Gene Oncology Analysis with STEM

• STEM analysis provides significance ranking of cluster gene profiles

• GO terms provide insight into function of gene clusters • Compare significant profiles for wild type and ΔGLN3

ANOVA WT ΔGLN3

p < 0.05 2378/6189 (38.42%) 1864/6189 (30.11%)

p < 0.01 1527/6189 (24.67%) 1008/6189 (16.29%)

p < 0.001 860/6189 (13.90%) 404/6189 (6.53%)

p < 0.0001 460/6189 (7.43%) 126/6189 (2.04%)

B-H p < 0.05 1656/6189 (26.76%) 913/6189 (14.75%)

Bonferroni p < 0.05 228/6189 (3.68%) 26/6189 (0.42%)

Proportion of Genes that Were Significant Followed Similar Patterns for the Wild Type and ΔGLN3 Mutant

The Wild Type and Mutant Had Four Significant Profiles in Common

Wild Type ΔGLN3

45 22 9 28 48 0 45 9 2 22 48 31

Profile #45 Had the Most Significant p-value for the Wild Type

Profile #45 had the Most Significant p-value for the ΔGLN3 Mutant

Profile 45 for Wild Type and ΔGLN3 Had a Pattern of Up-Regulated genes During the Significant Time points

Majority of GO list terms Dealt with Ribosome Biogenesis

WILD TYPE ΔGLN3

GO:0005730GO:0022613GO:0042254GO:0016072GO:0006364GO:0034660GO:0034470GO:0031981GO:0030684GO:0043233

GO:0022613 GO:0042254GO:0005730 GO:0034470 GO:0006364GO:0016072 GO:0030684 GO:0034660GO:0031981 GO:0016070

• All of the terms in both of our lists dealt with ribosome biogenesis, the process of making new ribosomes

• Cold temperatures stabilize RNA secondary structures, which causes ribosomes to be immobilized, slowing the normal functions of the cell (transcription, translation)

• To compensate, the cell makes more ribosomes

GRNSight Only DNA Binding Degree Had a Managable Gene Regulatory Network

• “DNA binding AND expression evidence” required there to be both DNA binding and expression evidence present, there was the least amount of connections shown by the GRNsight maps.

• There were not 15-30 transcription factors present

• “DNA binding PLUS expression evidence” only required there to be DNA binding or expression evidence present, there was the most amount of connections shown by the GRNsight maps.

GRNSight DNA only binding had the optimal number of edges

• “DNA binding PLUS expression evidence” 185 edges• “DNA binding AND expression evidence” 7 edges • “Only DNA binding evidence” 51 edges• The frequency distribution for “DNA binding AND

expression evidence” had very little data, so it is difficult to come to conclusions.

• The distribution for “DNA binding PLUS expression evidence” was roughly bell-shaped, with the most frequent number being 7.

• “DNA binding AND expression evidence” had a skewed right plot with it being most common to have a low number of connections.

GRNSight “Only DNA Binding” Had Optimal Edges a gene regulatory network

The desired amount of edges was 40-60, so the “Only DNA binding evidence” was the most optimal for further study

GLN3 (b=0 on the left, b=1 right)

The Wild Type and ΔGLN3 Strain Diverge Dramatically, But the Model Does Not and Therefore Does Not Fit the Individual Strain Data Well

Many Genes Showed Non-Zero Dynamics within the Model

CIN5, HMO1, INO4, MIG2, MSN4, PDR1, SFP1, SNF5, and YLR278C showed non-zero dynamics within the model

GRNsight Map for both Inputs Showed GLN3 Was Not Well Connected

Comparison of GRNsight outputs

b=1 on the left, b=0 on the right

Bar Graph comparing fixed and estimated weights

ControllerGeneA->TargetGeneB

CIN5-

>ASG1

CIN5-

>MIG

2

CIN5-

>SFP1

CIN5-

>YHP1

FKH2->ASG1

FKH2->SNF6

FKH2->YHP1

HMO1-

>CIN5

HMO1-

>FKH2

HMO1-

>MCM

1

HMO1-

>MSN4

HMO1-

>YLR27

8C

INO4-

>ACE2

INO4-

>SWI5

INO4-

>YLR27

8C

MCM

1->SNF6

MCM

1->YHP1

MIG

2->RIF

1

MSN2-

>CYC8

MSN2-

>MSN4

MSN2-

>SFP1

MSN2-

>YLR27

8C

PDR1->M

SN4

STB5->M

SN4

YHP1->GLN

3

ZAP1->ACE2

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

Weights fixed-b

Weights estimated-b

Production Rates of Fixed vs. Estimated b

ACE2

ASG1CIN

5

CYC8FKH2

GCR2GLN

3

HMO1

INO4

MCM

1M

IG2

MSN2

MSN4

PDR1RIF

1SFP1

SNF5SNF6

STB5SW

I5

YHP1

YLR27

8CYOX1

ZAP10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Fixed_b

Estimated_b

ΔGLN3 exhibits no meaningful control of cold shock response, other genes should be studied• GLN3 was only controlled by one gene and controlled no

genes• The fact that GLN3 did not activate or repress any other genes

means that there was no real effect when it was deleted• There does seem to be a relationship between goodness of fit

and noise in the data. • Any future direction should not involve GLN3, but rather a

more connected gene, such as CIN5, MSN4, or SPF1

Acknowledgments

• We would like to thank Professor Dahlquist and Fitzpatrick for their insight and assistance in creating our model

• We would also like to thank our wonderful classmates, for constructive partnerships and help, although we were examining different strains

Works CitedAlberts, B. (2002). Studying Gene Expression and Function. In Molecular biology of the cell (4th ed.). New York:

Garland Science.

Aguilera J, Randez-Gil F, Prieto JA. Cold response in Saccharomyces cerevisiae: new functions for old

mechanisms. FEMS Microbiol Rev 2007;31:327-41

Al-Fageeh, M. B., & Smales, C. M. (2006). Control and regulation of the cellular responses to cold shock: the

responses in yeast and mammalian systems. Biochemical Journal, 397(Pt 2), 247–259.

doi:10.1042/BJ20060166

Cherry JM, Hong EL, Amundsen C, Balakrishnan R, Binkley G, Chan ET, Christie KR, Costanzo MC, Dwight

SS, Engel SR, Fisk DG, Hirschman JE, Hitz BC, Karra K, Krieger CJ, Miyasato SR, Nash RS, Park J,

Skrzypek MS, Simison M, Weng S, Wong ED (2012) Saccharomyces Genome Database: the genomics

resource of budding yeast. Nucleic Acids Res. Jan;40(Database issue):D700-5.

Tai, S. L., Daran-Lapujade, P., Walsh, M. C., Pronk, J. T., & Daran, J.-M. (2007). Acclimation of Saccharomyces

cerevisiae to Low Temperature: A Chemostat-based Transcriptome Analysis. Molecular Biology of the Cell,

18(12), 5100–5112. doi:10.1091/mbc.E07-02-0131

Thieringer, H. A., Jones, P. G. and Inouye, M. (1998), Cold shock and adaptation. Bioessays, 20: 49–57. doi:

10.1002/(SICI)1521-1878(199801)20:1<49::AID-BIES8>3.0.CO;2-N

28

Creating a Gene Regulatory Network Comparing a Wild Type Strain with a Mutant ΔGLN3 Deletion in S....

Documents

Transcript of Creating a Gene Regulatory Network Comparing a Wild Type Strain with a Mutant ΔGLN3 Deletion in S....